Protein structure
The protein structure is divided in biochemistry at different structural levels. The classification into a hierarchy in primary structure ( amino acid sequence), secondary structure, tertiary structure and quaternary structure was first proposed in 1952 by Kaj Ulrik Linderstrøm long. In relation to the spatial arrangement of a protein, the term protein conformation is used synonymously. Changes in the spatial protein structure are called conformational changes.
The hierarchy of structural levels
In biochemistry, four hierarchically arranged structure levels can be distinguished in proteins:
- Primary structure - the amino acid sequence ( sequence of amino acids ) of the peptide chain.
- Secondary structure - the spatial structure of a local region in the protein ( eg, α -helix, β -sheet ).
- Tertiary structure - the spatial structure of a subunit.
- Quaternary structure - the three-dimensional structure of the entire protein complex with all sub-units.
Some proteins also arrange themselves in one of the quaternary structure beyond " superstructure " or " superstructure " to. This is exactly the same molecular structure predetermines how the other levels. Examples of superstructures are collagen in the collagen fibril, actin, myosin and titin in the sarcomere of Muskelfibrille and capsomeres in the capsid enveloped viruses.
Formation of a spatial structure
The process of three-dimensional space filling of a protein is partly spontaneously during translation, in part, the involvement of enzymes or chaperones is required. Also ligands affect protein structure, so that some proteins, depending on complexation with cofactors or substrates can assume different structures (see: conformational change ). This ability to change the spatial structure is needed for many enzyme activities.
Disturbances in the formation of a functional three-dimensional structure are referred to as a protein misfolding disease. An example is Huntington's disease. Diseases result from a malformation of the protein structure called prion diseases. BSE or Alzheimer's disease are examples of such illnesses. Also, diabetes mellitus type 2 is a protein misfolding disease, it is based on a misfolding of amylin.
Structure determination
To elucidate the spatial protein structure different experimental methods are available:
- In the crystal structure of an X-ray diffraction pattern of a protein crystal is - usually created by means of which it is then possible to calculate its three-dimensional structure. The preparation of the required single crystal is very difficult and for some proteins, not previously possible. Another problem with this method is that the structure of the proteins in the crystal does not necessarily correspond to the natural structure ( crystal packing ). For evaluable diffraction patterns of a minimum size of protein crystals is required. To obtain the amount of substance ebenötigte is often resorted to proteins produced by bacteria. These have sometimes not on the higher organisms existing in proteins post-translational modifications.
- By means of NMR spectroscopy, the structure of a protein in solution can be determined, which corresponds to the physiological ( "natural" ) conditions of the protein more. Since moving atoms of the protein in this state, there is no clearly defined structure. In order to obtain " clear " structure is usually averaged over the imaged structures. NMR spectroscopy can not yet be carried out for all types of protein. In particular, the size is a limiting factor. Proteins > 30 kDa can not be analyzed so far, since the NMR results are so complex that it is no unambiguous protein structure can be derived.
- The structure is dependent on a variety of physico-chemical conditions ( such as pH, temperature, salt content, the presence of other proteins). The Stokes radius of the native protein or a protein complex can be determined via a native- PAGE, size exclusion or isopycnic centrifugation. These two methods can be combined with a cross-linking or an alanine scan.
Structure prediction
The prediction of spatial protein structures achieved good results when there is already proteins with similar sequence and known structure. This allows the so-called homology modeling, wherein the new sequence to the sequence, whose structure is known, and thus ready to " fit " into the structure. This technique is similar to the sequence alignment.
More difficult is the prediction when no structures of similar amino acid sequences are known. The Levinthal paradox shows that the calculation of the lowest energy conformation due to the many possibilities is not feasible. In bioinformatics, great progress has been made in recent years and developed several methods of de novo or ab initio structure prediction. However, so far there is no reliable method for structure determination of proteins.
In order to compare new methods for structure prediction with each other, there are some years the CASP competition (critical assessment of techniques for protein structure prediction). In this competition, amino acid sequences of structures in which crystallographers working on, provided for the participants. Participants use their own methods to predict the structures. An evaluation team compares the predictions then with the experimentally determined structures.
Structure prediction was or is also the aim of several projects of distributed computing such as Rosetta @ home, POEM @ home, Predictor @ home, Folding @ home and Human Proteome Folding Project. The game Foldit was also influenced by the structure determination, the benefits of crowdsourcing advantage.